CHAPTER 3 – DATA AND METHODS
3.1 D ATA SET FOR MEASURING THE ASSOCIATION BETWEEN CRIMES AND SOCIAL INDICATORS IN E UROPE
This thesis uses a quantitative analysis method that permits the empirical testing of the proposed objectives; such testing will allow us to move beyond the narrative-review approach of the qualitative method. The available data only allows for a non-experimental approach.
3.1.1CRIMES FOR MEASURING THE ASSOCIATION BETWEEN CRIMES AND SOCIAL INDICATORS IN EUROPE
In order to measure the association between types of crimes and social indicators in Europe, seven categories of crimes were selected. Being aware of the limits of police statistics in accurately measuring crime levels (Aebi 2004, 2008, and 2010), victimization surveys are believed to be the most adaptable data source that can be used to achieve the first objective of this thesis because they present comparable data among different countries. The crimes selected for this study are: car theft, motorcycle theft, burglary, robbery, sexual offence, assault and threat, and intentional homicide.
These crimes have been selected by taking into account the data’s completeness and its analogous crime categories, as examined in the ESCCJ21
21 Chapter 5 presents the examination of crime trends. Data from the ESCCJ will be taken into account. The first part of the thesis (Chapter 4) is functional in relation to the second part (Chapter 5) because it permits the identification of social indicators that have a relationship with crime. These social indicators will then be used to explain trends.
. In particular, data related to car theft and motorcycle theft is related to the mean victimization rate for car and motorcycle owners. The data expresses the
‘victimization prevalence rates’, which refer to the percentage of the population 16 years of age or older who have been victimized in a specific crime in the course of the year 2004; this information is
58
organized by country (national one-year victimization prevalence rates).
Victimization surveys do not collect data related to homicides, so in order to measure the association between homicide and social indicators, the WHO data set (HFA-DB) was chosen for the following reasons:
- Comparable data: The WHO compiles annual transnational mortality data sets based on national mortality statistics so as to have comparable data on intentional homicides (homicides and intentional injuries) in Europe. The WHO measures homicides using the International Classification of Diseases codes and these attributes make the homicide statistics derived from death registration data more easily comparable across countries than the equivalent of those derived from criminal justice data (Small Arm Survey 2012).
- Accuracy and analogy: Some authors have argued that the WHO data is the most accurate extant dataset (LaFree 1999; Neumayer 2003; Small Arm Survey 2012) and it is one of the more often used data sources in transnational homicide studies (LaFree 1999). Moreover, Aebi (2012a, 2012b), is one researcher who has used the WHO dataset; he has greatly inspired my work and is another reason why the WHO dataset (HFA-DB) has been chosen for this work.
3.1.2SOCIAL INDICATORS FOR MEASURING THE ASSOCIATION BETWEEN CRIMES AND SOCIAL INDICATORS IN EUROPE
In order to measure the association between types of crimes and social indicators in Europe, this section explains the data gathering process for the social indicators and the operationalization of the macro theories (the modernization, civilization, and opportunity theories) into measurable units.
3.1.2.1 AN INTEGRATED APPROACH FOR CHOOSING SOCIAL INDICATORS FOR MEASURING THE ASSOCIATION BETWEEN CRIMES AND SOCIAL INDICATORS IN EUROPE
Lisa R. Muftić (2009) presented an interesting article that mentioned some of the most important criminologists who have proposed an applied and integrated approach to crime research over the last 20 years. She explained that criminology has been dominated by theories that are based on rigorously macro or micro level theoretical propositions. These theories, however, have generally failed in their ability to explain crime and criminality, so in response, some criminologists have begun to seek the integration of theoretical frameworks. It may be argued that almost all criminological theories are in
59
some form ‘integrated theories’, because they use consolidated concepts and propositions (Osgood 1998). In recent years, some attempts (Paternoster and Bachman 2001) have proposed integrated approaches that mix macro level and micro level theories so as to obtain an integrated theoretical model that maximizes the explained variance (Wellford 1989).
Akers and Sellers (2004) discussed how theoretical integration is a process in which two or more competing theories are combined to make a new theory, which provides a more comprehensive view of crime. Theories and integrated approaches use demographic, economic, and social indicators to examine associations between crime levels and risk factors. In this study, the integrated approach is the only one possible because an approach that is based purely on a single theory would, at best, produce partial results addressing a small portion of variance in crime.
3.1.2.2THEORETICAL FRAMEWORKS AND THE OPERATIONALIZATION PROBLEM FOR MEASURING THE ASSOCIATION BETWEEN CRIMES AND SOCIAL INDICATORS IN EUROPE
On the basis of suggestions offered in Section 3.1.2.1, a set of social indicators will be collected to explain crime levels in Europe and these indicators will be used to assess the generalization of three theoretical propositions (the modernization, civilization, and opportunity theories) across European crime levels. One of the main problems in the social sciences is the question of how to evolve from abstract concepts to concrete observations; the process that identifies phenomena in order to represent abstract concepts is called ‘operationalization’, which means to ‘quantify abstract concepts’. Three steps are required to do this: first, define concepts; second, identify properties;
third, choose social indicators. Below is a brief description of the steps.
a. Define concepts:
‘Development’ is a ‘trait d’union’ between the modernization and civilization theories; this is why it is used to operationalize the concepts of ‘civilization’ and ‘modernization’. Shelley (1981) used the The first step is giving a definition for the ‘civilization’, ‘modernization’, and
‘opportunity’ theories, since they are concepts that do not have absolute definitions. For example, Weisner and Abbott (1975) defined ‘modernization’ as the product of multiple experiences (e.g.
schools, farms, institutions, etc.), while Smith and Inkeles (1966) stated that ‘modernization’
generally means a national state characterized by a complex of traits, including urbanization, high levels of education, industrialization, and high rates of social mobility.
60
following definition of development, which has been provided by the United Nations Educational, Scientific, and Cultural Organization (UNESCO): ‘Development is an integral and interacting process, both requiring and precipitating far-reaching social, political, cultural and economic changes. It is by no means a[n] unlinear process that moves steadily and smoothly toward some predetermined set of models and values...[I]t is typically turbulent, often a downright disorderly and painful process.’
For ‘opportunity’, the most sensible definition is the one proposed by Cohen and Felson (1979), which explains the condition that may cause predatory acts as ‘the convergence in space and time of likely offenders, suitable targets, and the absence of capable guardians against crime’.
b. Identify properties:
The civilization and modernization theories may be operationalized through the concept of
‘development’, which is characterized by these properties: quality of life, education, economic development, health, technology, economic well-being, family relationships, and multiracialism.
The opportunity theory may be operationalized through the concept of ‘likely offenders, suitable targets, [and the] absence of capable guardians’ (Cohen and Felson 1979), which are components based on these properties: security, work, life balance, economic well-being, multiracialism, networks, and technology.
These definitions allow us to identify the concepts and then the principle concept can be broken down into its different properties, which can then be operationalized into statistical indicators.
c. Choose social indicators:
Currently, commissions (European or local) frequently have the task of selecting a set of social indicators to monitor social, economic, and demographic conditions concerning the entirety of The third step is choosing social indicators for each property. This choice is grounded in the available literature as well as data and the author of this thesis is conscious that many different choices could be made. An important factor to consider is that the concept changes according to historical periods, places, and cultures and therefore cannot simply be defined according to a theoretical format. However, the employed indicators are commonly used in similar research (BES, Istat-Cnel). The process and the choices made are simply one possibility in the universe of possibilities, and this particular route was taken while accounting for some guidelines regarding the representativeness of the social indicators.
61
Europe or individual countries. Scientific research in this field (Eurostat 2010; CNEL and Istat – BES) shows that at the moment, no single statistical indicator is capable of fully representing a society’s state of well-being, leaving us to refer to a range of measures. It is very common for different commissions to propose varied sets of social indicators.
Table 6 synthesizes the concepts needed to operationalize the properties and social indicators selected for civilization theory, modernization theory, and opportunity theory while table 7 shows the list of social indicators and defines them.
62
Table 6. Statistical indicators selected to operationalize civilization theory, modernization theory, and opportunity theory
Theory Concept to
operationalize
Sources for
operationalization Properties Social indicators
civilization theory and modernization
theory
development Eurostat 2010; CNEL and Istat - BES.
quality of life HDI
education school expectancy economic
development GDP per capita
health life expectancy at birth, infant mortality, healthy years technology science and technology economic
well-being severe material deprivation family relationships divorce
multiracialism acquisition of citizenship
opportunity theory
likely offenders, suitable targets,
absence of capable guardians
Cohen and Felson 1979; Eurostat 2010;; CNEL and
Istat - BES.
security part-time status, burglar alarms work and life
balance
long-term unemployment, resource productivity
economic
well-being severe material deprivation multiracialism acquisition of citizenship family network household type
technology science and technology
Below the social indicators selected are briefly explained and defined.
According to Sharpe and Smith (2005), the best known composite quality of life scale is the United Nations Development Program's Human Development Index (HDI). This index provides single values that measure the health and longevity, knowledge (literacy and school enrolment), and standards of living (GDP per capita) of a population. It permits us to compare development levels in different countries.
Concerning education, ‘school expectancy’ informs on the expected years of education over a lifetime, has been taken into account.
Concerning economic development, ‘GDP per capita’ is the gross domestic product at purchasing power parity per capita; this means that the GDP per capita is the gross domestic product divided by the mid-year population. As for the theme of ‘income index’, traditional scholarly works generally
63
focus on per capita income or GDP as the most crucial factor leading to democracy.
For health, ‘life expectancy’ informs us of the mean number of years that a newborn child can expect to live and replaces other social indicators that are commonly used like ‘mean or median age’. ‘Infant mortality’ measures the number of deaths per year of children younger than one year of age against the number of live births in that same year. The indicator of ‘healthy life years’ (HLY) measures the number of remaining years that a person of a specific age is expected to live without any severe or moderate health problems. In education, ‘school expectancy’ takes into account the expected years of education over a lifetime.
Data on marital status is a good indicator of family level disruption. The ‘divorce rate’ will also be taken into account here.
Immigration flows reveal information about the heterogeneity level that is present in a country; in particular, the social indicator ‘acquisition of citizenship’ depicts how immigrants have integrated into the host society. Cultural heterogeneity, which is a product of modern society, may lead to weaker communities (Howard et al. 2000; Sampson and Groves 1989).
Social indicators concerning economic structure are commonly used to examine the economic well-being of a country (Eurostat 2010). World Bank collects some measures of long-term structural change to evaluate the development process. There are social indicators for several relevant concepts, including economic growth and structure, government finance, labour force and employment, and money and prices. The social indicator selected is ‘severely materially deprived people’, which speaks to a population’s poverty level and economic inequality conditions. It permits the assessment of economic structure in macro units (i.e. by countries).
Moreover, ‘science and technology’ may be a representative factor of the technological degree that a country has achieved. These social indicators aptly describe recent changes in modern and civilized society. Some of the social indicators selected above are good predictors for opportunity theory as well. In particular, these indicators are: ‘acquisition of citizenship’, ‘severely materially deprived people’, and ‘science and technology’.
After that, this study considers two social indicators concerning the work and life balance that may be used to assess crime opportunities: ‘resource productivity’ and ‘long-term unemployment rate’.
‘Resource productivity’ is the GDP divided by the domestic material consumption (DMC), where DMC
64
measures the total amount of materials directly used by an economy. The ‘long-term unemployment rate’ is the share of people who have been unemployed for at least 12 months in comparison to the total number of active people in the labour market. This provides information on the volume of inoccupation. These social indicators reflect significant information on disposable income and allow us to represent the poverty element and the lack of resources that may lead to crime.
For the area of the ‘family network’, the social indicator ‘size of household’ gives information on the distribution of a population according to household types (e.g. single person, one adult older than 65 years, single person with dependent children, etc.). Two modalities have been chosen for this analysis:
‘single person with dependent children’ and ‘two adults with at least one aged 65 years or over’.
Finally, for the security section, the selected social indicators are ‘people employed part-time’ and
‘burglar alarms’. ‘People employed part-time’ represents the number of people employed part-time.
Eurostat advises that the distinction between full-time and part-time work is made on the basis of a spontaneous answer given by the respondent. It is impossible to establish a more exact distinction between the two types due to the variations in working hours between Member States and branches of different industries. After that, ‘burglar alarms’ demonstrate the security precautions that households have taken to protect their houses against burglary.
In summary, these are the social indicators that are used to test both the civilization and modernization theories: ‘HDI’, ‘life expectancy at birth’, ‘school expectancy’, ‘GDP per capita’, ‘infant mortality’, ‘divorce’, and ‘healthy years’. ‘ Acquisition of citizenship, ‘severe material deprivation’ and
‘science and technology’ will be used to test the civilization and modernization theories as well as the opportunity theory. The social indicators used to test only the opportunity theory are as follows:
‘resource productivity’, ‘long-term unemployment’, ‘household type’, and ‘burglar alarms’.
65
Table 7. Social indicators list
Social indicator Definition
human development index combination of social indicators regarding life expectancy, education, and income indices life expectancy at birth mean number of years that a newborn child can expect to live
school expectancy expected years of education over a lifetime, which is calculated by adding the single year enrolment rates for all ages
GDP per capita gross domestic product per capita. The volume index of GDP per capita in Purchasing Power Standards (PPS is expressed in relation to the European Union’s [EU27] average set to equal 100. If the index of a country is higher than 100, this country's level of GDP per head is higher than the EU average and vice versa.)
infant mortality per 1000 live births number of deaths of children under one year of age during a year compared to the number of live births in that same year. The value is expressed per 1000 live births.
divorce per 1000 persons number of divorces during the year compared to the average population in that year. The value is expressed per 1000 inhabitants.
healthy life years measures the number of remaining years that a person of a specific age is expected to live without any severe or moderate health problems.
acquisition of citizenship refers to grants of citizenship from the reporting country to people who have previously been citizens of another country or who have been stateless.
severely materially deprived people - % and per 1000 persons
covers indicators relating to economic strain, durability, housing, and the environment of the dwelling. Severely materially deprived people have living conditions severely constrained by a lack of resources. They experience at least four out of the nine following deprivation items:
they cannot afford i) to pay rent or utility bills, ii) to keep their home adequately warm, iii) to face unexpected expenses, iv) to eat meat, fish, or a protein equivalent on every second day, v) a week holiday away from home, vi) a car, vii) a washing machine, viii) a colour TV, or ix) a telephone.
science and technology tertiary graduates in science and technology per 1000 persons of a population aged 20-29 years
resource productivity GDP divided by domestic material consumption. DMC measures the total amount of materials directly used by an economy.
long-term unemployment rate the share of unemployed people for 12 months or more in the total number of active people in the labour market.
distribution of population by household types
distribution of population by household types (e.g. single person, one adult older than 65 years, single person with dependent children, etc.)
people employed part-time number of people employed part-time.
burglar alarms percentage of households with a burglar alarm Sources: Elaborated from Eurostat, EU-SILC, LFS, HETUS, and WHO.
In the interest of transparency, Table 8 shows the data sources that were used to collect the social indicators reported in Table 7.
66
Table 8. Data sources for selected variables
DATABASE VARIABLES
Selected Eurostat indicators life expectancy at birth, infant mortality per 1000 live births, healthy life years, divorce per 1000 persons, science and technology , acquisition of citizenship, school expectancy
The European Union Statistics on Income and Living Conditions (EU-SILC)
GDP per capita, resource productivity, severely materially deprived people - % of 1000 persons
The Employment and Unemployment (Labour Force Survey)
people employed part-time, very long-term unemployment rate Harmonised European Time Use
Surveys (HETUS 2000)
distribution of population by household types European Health for All Database
(HFA-DB)
human development index ICVS – EU ICS burglar alarms
3.1.2.3SPATIAL AND TEMPORAL DIMENSIONS FOR MEASURING THE ASSOCIATION BETWEEN CRIMES AND SOCIAL INDICATORS IN EUROPE
This section summarizes the spatial and temporal coordinates used in measuring the association between types of crimes and social indicators in Europe. For the WHO, the HFA-DB’s most recent data is related to the year 2011 for a sample of 55 European and non-European countries, while the most recent data from ICVS and the EU ICS are related to the years 2004-2005 for a sample of 22 European countries. The EU ICS collects data from Austria, Belgium, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Luxembourg, the Netherlands, Poland, Portugal, Spain, Sweden, and the United Kingdom, while the ICVS collects data for a larger set of European countries (the same countries as the EU ICS, as well as Bulgaria, Iceland, Norway, and Switzerland) and some non-European countries.
For homicides, a broader set of countries (30) has been taken into account and more recent data is available. The selected countries for this crime are: Austria, Belgium, Bulgaria, Cyprus, the Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Iceland, Italy, Latvia, Lithuania, Luxemburg, Malta, the Netherlands, Norway, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, Sweden, Switzerland, and the United Kingdom. Altogether, these countries account for the
67
EU27 and three of the four States that belong to the EFTA22: Iceland, Norway, and Switzerland23
Table 9 shows European crime data that has been collected by the WHO and the ICVS/EU ICS in 2004/2005. The table also indicates which countries belong to the EU27. For homicides, the regions included are the EU27, as well as Iceland, Norway, and Switzerland, while the victimization surveys of crimes include 19 of the EU27 countries, as well as Iceland, Norway, and Switzerland.
. Without taking into consideration the existence of newer accessible data (from the year 2011), the year 2004 has been selected for temporal homogeneity within the study.
22 The European Free Trade Association (EFTA) is a free trade organization between four European countries that operates in parallel with—and is linked to—the European Union.
23 There is no data from the WHO for homicide in Liechtenstein.
68
Table 9. WHO and ICVS/EU ICS crime data for European countries, years 2004-2005 Country WHO homicide ICVS and EU ICS crimes EU27
Austria x x x
Belgium x x x
Bulgaria x x x
Cyprus x x
the Czech Republic x x
Denmark x x x
Estonia x x x
Finland x x x
France x x x
Germany x x x
Greece x x x
Hungary x x x
Iceland x x
Ireland x x x
Italy x x x
Latvia x x
Lithuania x x
Luxembourg x x x
Malta x x
the Netherlands x x x
Norway x x
Poland x x x
Portugal x x x
Romania x x
Slovakia x x
Slovenia x x
Spain x x x
Sweden x x x
Switzerland x x
the United
Kingdom x x x
3.1.2.4DATA MATRICES FOR MEASURING THE ASSOCIATIONBETWEEN CRIMES AND SOCIAL INDICATORS IN EUROPE
Having sufficient data is not enough to apply statistical elaboration; the information needs to be organized into a format that permits further elaboration. In general, to combine statistical data into a matrix, it is necessary to identify the ‘object’ (x) component and the ‘attribution’ (y) component.
Usually, the objects are displayed in rows, while the attributions are in columns. In this case, the table
69
is set up to present ‘individual features’, where the objects are the statistical units and the attributions are the features (variables) (Bolasco 1999).
The first data set (Appendix A) collects data for crimes from the ICVS and social indicators in 22 European countries while the second data set (Appendix B) collects data for completed homicide from the HFA-DB (WHO) and social indicators in 30 European countries. In summary, both datasets (Appendices A and B) collect figures respecting these conditions:
- Time: 2004-2005
- Space: 22 European countries for the ICVS offences (Appendix A) and 30 European countries for the WHO homicides (Appendix B)
- Variables: 15 social indicators and seven types of crimes
The temporal interval choice may be unpopular because it does not account for the most recently available data but it does have the following advantages for crime data. It:
- Has data for a broad sample of countries
- Avoids the utilization of different data sources, which would imply different definitions in terms of rules, units, time, etc.
- Uses ‘old’ data, thereby reducing the amount of missing data. In the case of missing data, however, the gap has been populated with the previous year’s value, if it was available.
The social indicators data is complete and seldom yields missing data. For household information, the missing values have been refilled with values recorded in previous years; otherwise it would not have been possible to obtain a representative variable. The spatial choice to take into account two different country samples was fuelled by the desire to test data in a broad sample; in fact, crimes collected from victimization surveys were only available for 22 regions, while the WHO homicide data set was larger. This decision permitted the testing of statistical elaborations in a wider sample of European countries (30).